Arcane: An Assertion Reduction Framework through Semantic Clustering and MCTS-Guided Rule Exploring

Arcane is a new assertion reduction framework that uses semantic clustering and Monte Carlo Tree Search to eliminate redundant assertions in hardware verification, achieving up to 76.2% reduction in assertion count while maintaining full formal coverage and enabling 2.6x to 6.1x simulation speedups.

Arcane addresses a critical bottleneck in hardware verification workflows where LLM-based assertion generation produces excessive redundant checks that bloat simulation environments and slow development cycles. The framework applies a two-tier semantic clustering approach to categorize assertions accurately, then deploys Monte Carlo Tree Search optimization to identify and eliminate duplicates while preserving verification integrity. This combination of techniques demonstrates how structured algorithms can complement automated generation tools by filtering their output intelligently.

The redundancy problem in assertion-based verification has grown with the rise of AI-assisted design tools, which tend to over-generate safeguards without understanding context-specific optimization. Hardware teams face increasing pressure to verify complex designs quickly, making simulation efficiency a competitive advantage. The documented 2.6x to 6.1x speedup translates directly to shorter iteration cycles, faster time-to-market, and reduced computational infrastructure costs.

For hardware development teams and EDA tool vendors, Arcane offers a practical post-processing solution that works with existing assertion generation pipelines. The framework's ability to maintain formal coverage and mutation-detection capability while cutting assertion counts substantially reduces simulation bottlenecks without sacrificing verification quality. This positions Arcane as a valuable complement to LLM-based design automation tools. Industry adoption depends on integration ease with existing verification flows and whether similar principles extend to other redundancy problems in hardware design toolchains.

• →Arcane reduces redundant assertions by up to 76.2% while preserving formal coverage and mutation detection
• →Two-tier semantic clustering combined with MCTS enables efficient assertion optimization without losing verification integrity
• →Simulation speedups of 2.6x to 6.1x directly accelerate hardware design cycles and reduce computational costs
• →The framework addresses a growing bottleneck created by over-generation in LLM-based assertion tools
• →Results validate on Assertionbench, demonstrating real-world applicability in production hardware verification